View metadata, citation and similar papers at core.ac.uk
brought to you by
CORE
provided by Saga University Institutional Repository
LOWLAND TECHNOLOGY INTERNATIONAL Vol. 14, No. 2,70-83, December 2012
International Association of Lowland Technology (IALT), ISSN 1344-9656
CURRENT PRACTICE ON FOUNDATION DESIGN OF HIGH-RISE BUILDINGS IN
BANGKOK, THAILAND
K. Amornfa 1, N. Phienwej 2 and P. Kitpayuck 3
ABSTRACT: Assessment was made on the current practice on foundation design of high-rise buildings in Bangkok,
Thailand to explore rooms for improvement. An interview survey revealed that the current design practice was
dominated by structural engineers. They commonly used the conventional method of analysis, namely the combined
stress equation, as well as the plate on springs analysis. The finding from the survey study indicates the current design
practice does not encourage an optimal design outcome in term of cost effectiveness. The second part of the study is to
explore the benefit in adopting the piled raft foundation design concept. A comparative study on the results of the 3dimensional finite element (3D FEM) analysis and various analysis methods currently used. The results show that the
plate on pile springs method which neglects pile-pile and raft-pile interaction give results significantly different from
that of the 3D FEM. The 3D FEM shows that only about 70-80% of total building loads are carried by piles when raft is
placed in the stiff clay layer. The number of piles in the piled raft foundation can be significantly reduced particularly if
the true piled raft foundation concept is adopted, while the foundation settlement only increases slightly.
Keywords: Foundation design, high-rise building, piled raft foundation, three-dimensional finite element method.
INTRODUCTION
Bangkok, the capital city of Thailand, has been
witnessing a rapid increase in number of tall buildings
construction for the past two decades. Because the city is
situated in a vast deltaic plain underlain by a series of
thick subsoil layers, the buildings need to be founded on
piled foundations to transfer load to soil strata at depth.
Typically, tall building in Bangkok required basements
for purpose of car park space which is stipulated by law.
Therefore, piled foundations start at depths of 10-20 m
below the ground surface; and often large numbers of
piles located at close spacing are utilized owing to large
building loads on relatively small land areas. The design
has traditionally followed “piled foundation” concept for
which all building loads are to be carried by piles. This
traditional foundation design concept of Bangkok was
justified when considering the preference of minimizing
building settlement and potential gap formation
underneath the pile cap in long term resulted from land
subsidence phenomenon caused by deep well pumping
for ground water supply (Phienwej et al. 2006). However,
at the turn of the century, the mitigation measures of
Bangkok land subsidence have been much improved,
consequently the annual rate of subsidence in the inner
city areas where there will be more tall building
development has been much reduced or even ceases.
Thus, it may be no longer a wise design approach by
ignoring the soil bearing resistance below the large-area
pile cap in the design.
Nowadays, the “piled raft foundation” concept has
been increasingly advocated and adopted in design of tall
buildings in many parts of the world (Poulos and Davis
1980; Randolph 1983; Yamashita et al. 1994;
Kachzenbach et al. 2000; Poulos 2001; and de Sanctis
and Mandolini 2006) because it has a potential costsaving and a better control of differential settlement. The
design makes use of the soil bearing resistance below the
raft together with the load carrying capacity of piles. To
ensure continuity of the soil bearing resistance to the raft,
the piles are designed at a sufficiently lower safety factor
than that normally adopted in the traditional piled
foundation design so that settlement of the piles are large
enough to ensure consistent contact between soil and raft.
However, the magnitude of settlement must be within a
tolerable limit of building function and structural safety.
With the advance in numerical methodology and tools,
the analysis of a piled raft foundation, that was once a
complex problem and could not be simply adopted in
1
School of Engineering and Technology, Asian Institute of Technology, THAILAND
School of Engineering and Technology, Asian Institute of Technology, THAILAND
3
Engineering Division, Seafco Public Company Limited, THAILAND
Note: Discussion on this paper is open until June 2013
2
－ 70 －
Current practice on foundation design of high-rise buildings in Bangkok, Thailand
design practice, now becomes available. Thus the option
of piled raft foundation design should be explored.
It is the objective of this study to investigate the
benefit of adopting the piled raft foundation concept for
foundation design of tall buildings in Bangkok. In the
study, firstly the current design practice of piled
foundation of tall buildings was reviewed by means of
interviews with a group of sampled design firms. In
investigation of the application of piled raft foundation
in Bangkok area, a 3D FEM analysis using PLAXIS 3D
Foundation software (Brinkgreve and Broere 2004) was
conducted. PLAXIS 3D Foundation is a rigorous
geotechnical code that is capable of analyzing directly
soil-structure interactions. A comparative study was then
made on the result of the 3D FEM analysis and that
obtained from the simplified FEM structural analysis, i.e.
Plate on springs modeling, commonly adopted for piled
foundation design in Bangkok.
SUBSOILS IN BANGKOK
Bangkok is located on the lower Chao Phraya plain,
which is made of a very thick deposit of marine and
alluvial soils. The subsoil layering is relatively uniform
over the entire city area and consists of a 10-15 m thick
layer of soft marine clay that is underlain by a very thick
series of alternating stiff to hard clay and sand and
gravelly sand. The typical subsoil is shown in Fig. 1.
Bed rocks are only found at great depths.
Because of the existence of the thick soft clay layer
at the ground surface, all buildings in Bangkok are
founded on piles to transfer load to soil layers at depth.
Various pile depths are used depending on type and size
of structures as well as the era of construction as also
shown in Fig. 1. Short friction piles (6 to 12-m depth)
floating in the soft clay layer were commonly used for
old structures. For later period, driven concrete piles or
bored piles with tips extending to the first sand layer at
18 to 30-m depth are commonly employed for light to
medium sized structures. For tall buildings, deep largediameter bored piles with tips extended to the second
sand layer (40 to 60-m depth) are usually used. Recently,
barrette piles (size 1 m x 2 m to 1.5 m x 3 m) constructed
by means of the diaphragm wall technique are also
adopted in a few tall building projects where the required
working load per pile exceeded 20,000 kN.
Nowadays, there are more than 4,500 buildings
exceeding 23-m height and 14 buildings exceeding 200m height in Bangkok. The number of high-rise buildings
has been increasing dramatically in the inner city area.
The list of the tallest group of building as of 2012,
excluding those under construction, is shown in Table 1.
These buildings were mostly designed to have concrete
mat slabs to transfer superstructure load to bored piles.
The thickness of the slab ranges from 2.5 to 3.5 m, but in
some building it reached 5.0 m. From the survey, it was
estimated that the cost of piled foundation of these tall
buildings was 5-10% of total construction cost.
CURRENT DESIGN PRACTICE
Design Group and Approach from Interview and
Literature Review
Currently, there are about 30-40 tall building design
companies working in Bangkok market. A systematic
interview was made with 10 chief/senior engineers of
about 30% of total design companies. They have been
working on foundation design for high-rise buildings in
Bangkok. Interview questions cover issues on interaction
between structural engineers and geotechnical engineers,
preliminary design, and detail design.
Table 1 List of tallest building with height exceeding
200 meters in Bangkok 2012
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Fig. 1 Typical Bangkok subsoil profile and range of
pile length (After Phienwej et al. (2006) and
Thasnanipan et al. (2006))
－ 71 －
Building
Baiyoke Tower II
The River Tower A
Meritus Suites State Tower
Centara Grand Hotel
The Met
Empire Tower 1
Jewelry Trade Center
Amanta Lumpini
China Resources Tower
Thai Farmer Bank
Headquarters
Central World Tower
The Pano
Terminal 21
Q-House Lumpini
Height
304
266
247
235
228
227
221
212
210
208
Storey
85
73
63
57
69
62
59
61
53
42
204
202
202
202
45
55
42
39
Amornfa, et al.
The survey was conducted based on intensive faceto-face interviews with a small number of respondents to
explore their current practice. Determination of
qualitative sample size relies on the concept of
“saturation point” or the point at which no new
information or themes are observed in the survey. Based
on this survey data, it found that saturation occurred
within the first 10 interviews. Variability of the current
practice within 10 respondents followed similar patterns.
On the basis of the survey result, the current practice
is assessed so that recommendations for improvement
can be made to achieve a better engineering practice and
economical outcome in future project.
Interview group
Typically the structural engineers are generally
charged with analysis superstructures, and often they
would handle the geotechnical design part as well. That
is to say for some design projects, there was no
engagement of geotechnical specialists in the foundation
design of the building. Thus for ease of reference in this
study, the author classifies 2 types of design group, i.e.
(i) SI Group - structural engineers solely responsible for
the design, and (ii) SGI Group - both structural engineers
and geotechnical specialist taking part in the design.
Design approach and procedure
The step in foundation design of all project starts
from soil investigation. This task is handled by a
qualified soil investigation company who has
geotechnical engineer to interpret subsoil stratigraphy
from borehole data and then suggest load capacity of
single piles, usually by the static equation method.
The next step of work after obtaining soil
investigation report is foundation design of the building
which is handled by building design companies.
According to the survey, the delegation of task in this
design step as commonly practiced for Bangkok case can
be summarized below for two groups of design
companies, i.e. SGI and SI Groups.
a) SGI group (70% of total interviewees): For this
group, the design was done through cooperation between
structural and geotechnical engineers. However, most
companies assigned structural engineer as main designer
to handle most parts of foundation design tasks,
including pile group analysis and concrete mat slab
structural design (57% of this group), as shown the work
flow chart of the design work and responsibility in Fig. 2.
b) SI group (30% of total interviewees): Design
companies engaged only structural engineers to do all
the design tasks of the building project including piled
foundation design. The chosen allowable load carrying
capacity of the pile is taken directly from the
Fig. 2 Work flow chart of the design work in SGI
group
recommendation in the soil investigation report. In this
design group, settlement analysis was neglected. Most
structural designers take it for grant that settlement of
building would be insignificant if pile tips were placed in
a layer of dense sand. The criterion on determination of
allowable load capacity of piles from static pile load
tests also specifies a pile head settlement limit at a small
value, typically 10 mm.
Design Calculation Steps
The design of foundation of tall buildings can be
divided into the following steps:
- Initial design step: Determination of column loads,
subsoil investigation, and single pile capacity analysis.
- Detailed design step: Design of pile group including
raft sizing and number of piles. This step also includes
structural design step and analysis for settlement check.
Single pile capacity analysis
In first step, non-factored column loads from the
superstructure were determined by structural engineers.
In next step, the suggested values of allowable capacity
of single piles at a site are normally given from static
equation analysis based upon soil parameters from a
investigation program made. Once pile size and tip depth
selection is made the adopted value is verified by
performing static pile load tests on test piles.
As the scope of this study is on high-rise buildings,
the current practice on deep bored piles with tips in the
second sand layer is focused. On the basis of past
experiences on design and pile load test results of large
diameter bored piles, the range of pile capacity can be
summarized in Fig. 3.
The survey shows that there were 2 approaches
applied for analysis of single pile capacity: 1) the static
equation method to determine ultimate load carrying
capacity (all interviewees used this method); and also 2)
finite element method (FEM) (10% of the interviewees)
－ 72 －
Current practice on foundation design of high-rise buildings in Bangkok, Thailand
summarized in Figs. 4-6. The suggested K tan value
significantly increases over the year, while for Nq and ,
a clear trend of change is not observed.
Earlier, bored pile construction in Bangkok was
made by means of rotary drilling with auger and bucket
method using bentonite slurry as supporting fluid for
borehole. But at present, the construction techniques
were improved to increase the pile load carrying capacity.
The technique includes cement grouting of soil at pile
base and side (Submaneewong 1999) and the addition of
polymer fluid in the bentonite slurry (Boonyarak 2002).
Since bentonite slurry tends to create filter cake on the
borehole wall thus can significantly reduce skin friction
of pile in sand layers. Adding a small amount of polymer
in bentonite slurry helps minimize thickness of filter
Fig. 3 Typical range of allowable capacity of bored
pile in Bangkok
Source: SEAFCO public company limited (2011)
In the static equation, end-bearing capacity is
Qp 
Ap q p 
A p cN c  q ' N q 
(1)
where Ap is area of pile tip, qp is unit end-bearing
resistance, c is cohesion of soil supporting the pile tip, q’
is effective vertical stress at the level of the pile tip, and
Nc, Nq are the bearing capacity factors.
The skin friction capacity is

Qf
  pLf 
(2)
where p is perimeter of pile section, L is incremental
pile length over which p and f are taken constant, and f is
unit friction resistance at any depth.
The unit friction, f , is
f  ( K tan  ) v '
f  c
for sandy soil
(3)
for clayey soil
(4)
where v’ is effective vertical stress at depth under
consideration, K is earth pressure coefficient,  is soilpile friction angle, and  is empirical adhesion factor.
The static equation method is theoretically sound as
it is described in all foundation engineering books.
However, the calculation depends on the accuracy of soil
parameters determination (e.g. Nq, , and K tan), which
are affected by many factors.
At the early period of bored piles usage in Bangkok,
the safe load adopted by the designer was rather
conservative. But with the gain of experience in
construction technique and quality load test data and
analysis (Ng 1983; Pimpasugdi 1989; Oonchittikul 1990;
Soontornsiri 1995; Submaneewong 1999; Boonyarak
2002; Chaiyawan 2006), nowadays much higher
capacity values can be used. The derived pile design
parameters from various studies through time are
Fig. 4 Relation between Nq and  for normal bored
piles in Bangkok
Fig. 5 Relation between  and Su for normal bored
piles in Bangkok
Fig. 6 Relation between K tan and
bored piles in Bangkok
－ 73 －
 for normal
Amornfa, et al.
cake. The improvement in load carrying capacity of
bored piles in Bangkok subsoils by these techniques can
be seen in the back-calculated parameters (Figs. 7 and 8).
Fig. 7 Nq for based grouted bored piles (After
Submaneewong (1999))
Fig. 8 Ktan for bored piles using polymer or based
grouting (After Submaneewong (1999) and
Boonyarak (2002))
Only 10% of all interviewees stated that they also
conducted FEM to determine pile capacity to supplement
the static equation analysis. The FEM has advantage that
it also yields load settlement behavior of the pile that
may be used to compare with a static pile load test result.
A few numbers of researches were made on analysis of
single pile in Bangkok subsoils using the FEM (e.g. de
Silva 1980; Fernando 1992). However, its application in
the real design practice is still not widely adopted. The
main reasons are non-engagement of geotechnical
specialists in the foundation design and problems in
selection of input soil parameters for the analysis. This
aspect should be improved in foundation design practice
in Bangkok so that a more powerful tool for foundation
design is efficiently exploited.
Factor of safety on pile capacity
To ensure safety on axial load capacity of each pile,
the design commonly adopt a safety factor (FS) of 2.5
for determination of allowable load from the total
ultimate capacity determined from the static calculation
method and static pile load test results. The value is
specified in the code of Building Control Act 1979. For
some exceptional case, the FS was reduced to 2.0 when
static pile load tests were available. The FS of 2.0 is also
recommended by International Building Code 2006,
Uniform Building Code 1997, the Engineering Manual
of U.S. Army Corps of Engineers, NAVFAC DM-7.2,
and AASHTO Standard specification for pile capacity
with verification from pile load test results. Thus, it can
be said that the commonly used FS of 2.5 in the building
design in Bangkok is on a conservative side. The use of
reduced FS can, consequently, reduce cost of building
foundation.
Pile group capacity
All design engineers determined pile group capacity
using the conventional method. In addition, 60% of them
did a second step check by adopting the plate on springs
analysis to determine more precise requirement of pile
capacity and number of piles with the understanding that
the method would improve the reliability of the design.
In the conventional method, the pile group capacity is
determined from the sum of single pile capacity of all
piles in the group with a group efficiency correction. The
most commonly used correction method is the ConverseLabarre’s. Normally, group efficiency correction is
necessary for friction piles, not end bearing piles. Per
ASCE guideline, the group efficiency is necessary for
friction piles in cohesive soils or piles in granular soils
with pile spacing closer than 3 times pile diameter.
Normally, all piles were designed as equally loaded and
the centroid of all piles aligns with the resultant vertical
superstructure column loads. In case that they can not be
aligned, the load distribution on the piles in the group is
determined from the combined stress equation. The
equation is

Qm
Q M yx Mx y


n  x2  y 2
(5)
where Qm is Load on any given pile, Q is Total vertical
load, n is Number of piles in the group, Mx, My are
Moment about x and y axes, respectively, and x, y are
Distance from x and y axes to any given pile
When the plate on springs analysis is used, piles in
the group will support unequal loads even when the
centroid of pile group is aligned with resultant of column
loads. This is the main different from the assumption in
the conventional method. The plate on springs method
does consider the effect of raft stiffness on the pile load
distribution in the group. Axial load on each pile from
the analysis is compared with the allowable single pile
capacity. If it exceeds the capacity, position of piles will
be adjusted or additional piles, normally 10-15% of
conventional pile number, are added. This design
practice may be considered too conservative. Because,
each of the piles in the group are designed to satisfy the
FS of a single pile even though it acts as a member of the
－ 74 －
Current practice on foundation design of high-rise buildings in Bangkok, Thailand
group. This design practice then results in a FS of the
pile group much higher than the required FS of single
pile acting alone.
In the plate on springs analysis the key input
parameter is the spring stiffness representing a pile for
which the selection of correct values for the analysis has
always been a question. Normally, the value is
preferably determined from the load settlement curve
from static pile load tests. Few other methods also used
in Bangkok practice, when pile load tests are not
available, are the empirical estimate of kp = 2EA/L (E =
young modulus of pile, A = cross sectional area of pile,
and L = pile length) and FEM via PLAXIS software to
simulate pile load settlement curve.
Based on an analysis on load settlement curves of
237 static pile load tests of piles in Bangkok subsoils,
Kiattivisanchai (2001) summarized that kp values of
bored piles are in the range of 0.5EA/L to 4EA/L with the
mean value of 2EA/L as shown in Fig. 9.
The weakness of the plate on springs type of analysis
for pile group foundation is that it can not directly
consider the interaction effect between piles in the group.
Owing to the pile interaction effect, edge piles will
behave stiffer than the inner piles even though they are
of the same size. The method to consider this effect may
follow the superposition principle given in a number of
literatures, e.g. Poulos and Davis (1980), Randolph
(1983), and Kitiyodom and Matsumoto (2002).
Alternatively the approximation method is by doubling
the exterior edge spring as proposed by Bowles (1986)
may also be used. However, for a large pile group this
spring stiffness correction method is not simple and
reliable to apply.
Fig. 9 The relation between axial stiffness and
vertical stiffness of bored piles (After Kiattivisanchai
(2001))
Structural design of foundation
In accordance with the methods employed for pile
group analysis, the bending moment and shear force in
the raft for structural design are determined from the
conventional calculation method and plate on pile
springs method. In Bangkok practice, 80% of the
interviewees stated that they used the latter method.
Settlement analysis
From the survey, 70% of total interviewees checked
the settlement condition. The remaining 30% did not
bother to check because they take it for granted the
settlement is small once they specifies depth of tip of
bored piles in the second sand layer. As the matter of
fact, all interviewees stated that the design should
follows the criteria that pile tips shall only be placed in
the sand layer, not in the stiff to hard clay layer. For
differential settlement control, they adopted the design
criteria that all piles used in the building must have the
same pile tip depths. In addition, separation/movement
joints are provided along the interface between the
podium and tower zones of a building to prevent
potential damages from differential settlement.
The equivalent raft method suggested by Tomlinson
and the one-dimensional consolidation method were
popularly used for calculation of building settlement in
Bangkok (71.4% of interviewees who do settlement
analysis). The remaining adopted 2D FEM analysis such
as PLAXIS. In the FEM analysis, undrained analysis was
used for prediction of short-term settlement and drained
analysis was used for long-term settlement. The plate on
pile springs analysis also give settlement of foundation
but it was considered as the short term settlement. Some
designer used it as a basis to check differential
settlement of the foundation. Currently, none of the
designers adopted 3D FEM for settlement analysis of the
high-rise buildings.
In the current design practice, it is considered that the
settlement prediction of high-rise buildings in Bangkok
subsoils is not a simple task to perform thus the results
of prediction by all methods currently used are
questionable. One reason is that there have been on few
cases of systematic monitoring of building settlements
from the start of the construction to the completion. As
an example, record of settlement data of one high-rise
building in Bangkok is shown herein. The building had
35 storeys with raft size 81 m x 47 m. Its foundation is a
rectangular raft of 2.5-m thickness resting on 1-m
diameter bored piles. Pile tip depth is 40 m below
ground surface in the second sand layer. The numbers of
piles under the tower is 299. The design load per pile
was 4100 kN/pile. At the construction time of 15th floor,
which 43% of completed floor, the settlements recorded
from both towers were about 4-5 mm. These settlements
were almost half of predicted immediate settlements of
13.2 mm. Moreover predicted consolidation settlements
will reach to 160 mm for this tower.
－ 75 －
Amornfa, et al.
Bangkok. The 3D FEM via PLAXIS 3D Foundation was
adopted for analysis of a high-rise building in Bangkok
Subsoils. The building under considered is a 50-storey
tall building with basements. The raft depth is varied
from 0 to 16 m below the ground level. The typical
Bangkok subsoil conditions as used are shown in Table 2.
For the first case, a 1-m thick raft with 9 bored piles
of 1-m diameter with tips at 46-m depth is considered.
The engineering properties of piles and raft are shown in
Table 3. The assumed vertical load conditions acting on
the raft is shown in Fig. 11. Because there was a
basement, the load was set to apply along the basement
wall as a line load and center point load for the lift core.
APPLICATION OF PILED RAFT FOUNDATION
CONCEPT
In piled raft foundation design concept, some portion
of the total building load is transferred directly from the
raft to the soil, therefore load carried by piles is reduced
and the number of piles could be minimized. The design
of a piled raft foundation requires an understanding of
complex soil-structure interaction which effects to the
load-settlement behavior (Katzenbach et al. 2000), Fig.
10. To take into account this complicated interaction and
to assess settlement of piled raft, 3D FEM analysis is
essentially which nowadays become readily available.
Fig. 10 Soil-structure interaction effects for piled
raft foundation (Katezenbach et al. 2000)
Fig. 11 Geometry of piled raft foundation at various
depths
In real practice, the design concept has been applied
in several countries, e.g. Germany, England, Japan,
Singapore, etc. For Bangkok, the design strategy
following this concept is not yet well developed. This
study attempts to make an initial investigation on the
potential benefit of application of piled raft foundation
design concept of high-rise buildings in Bangkok.
In the FEM analysis, 15-node wedge elements are
used to model soils and volume piles as shown in Fig. 12.
The total number of elements was 16,603 for case of raft
at ground level. For the case of raft depth below ground
surface, only soils below the raft level were modeled as
finite element mesh. Soils above the raft level were
simplified as a surcharge. Thin plate elements were used
to simulate the raft. The plate elements (6-node
triangular elements) and one side of soil elements (15node wedge elements) were connected. The connection
Numerical Modeling of Piled Raft Foundation Analysis
A parametric study was made together with a
comparative study with design methods currently used in
Table 2 Bangkok subsoils and material properties (After Suwanwiwattana (2002))
Material
Depth
Young’s modulus
Poisson ratio
Unit weight
Friction Angle
Cohesion
Interface parameters
Soft clay
Stiff clay
1st sand
Hard clay
2nd sand
0-15
15-25
25-35
35-45
45-55
3000
40500
80000
150000
200000
0.495
0.495
0.3
0.495
0.25
15.2
18.4
19.4
19.8
20.1
35.8
36.2
20
90
300
-
0.9
0.5
0.6
0.3
0.6
(m)
Eu (kN/m2)

(kN/m3)
cu (kN/m2)
(degree)
Table 3 Engineering properties of piles and raft
Material
Bored pile
Raft
Depth
(m)
Size
Young’s modulus Poisson ratio Unit weight Material model
- Tip at 46 m
- 1-m diameter
- Top vary from 0-16 m - Length vary from 30-46 m
- 1-m thick
Vary from 0 to 16 m
- 9 m x 9 m size
－ 76 －
Eu (kN/m2)

(kN/m3)
26.0 x 106
0.2
24
Linear elastic
28.0 x 106
0.2
0
Linear elastic
Current practice on foundation design of high-rise buildings in Bangkok, Thailand
Fig. 13 Relation of load shared and raft depth in
Bangkok subsoils
Various Methods of Design Analysis
Fig. 12 Finite element mesh with raft and piles in
PLAXIS 3D Foundation
was not rigid connection. Interface behavior between
plate elements and soil elements are taken care by Rinter.
The Mohr-Coulomb model (linear elasto-plastic) was
used to model soil properties. The 16-node interface
elements were introduced on the surface of the pile. The
interface elements had zero thickness and their strengths
were scaled down from the strength of surrounding soils
by interface parameters which were equivalent to
adhesion factors for determination of the skin frictions of
soils. Only short term behavior of the piled raft
foundation was investigated, thus the total stress
undrained strength parameters of clay layers were used
in the analysis. Consolidation process during
construction was not considered in 3D FEM. Since pile
tip in this case of study is resting on sand layer,
consolidation during construction may not be so
significant as in the case of pile tip resting on clay layer.
A comparative study is made by performing the
analysis using the common methods of design analysis
for piled foundation currently used in Bangkok with the
rigorous 3D FEM piled raft analysis with PLAXIS 3D
Foundation. A case of a raft on 25 bored piles is
considered in this case. The raft is 1-m thick and placed
in the stiff clay layer at 16-m depth. The piles are of the
same conditions as those used in the analysis above. The
ultimate capacity of each pile from the static equation
analysis is 11.25 MN. Two conditions of building loads
acting on the raft is considered namely, i) uniform
loading over the entire raft area and ii) a line load along
basement wall and central point load at lift core, Fig. 14.
Load Shared between Piles and Raft
Fig. 13 shows load shared by piles and raft for
different of raft depth below the ground surface. The
analysis shows that when the raft level is placed down to
the stiff clay layer, the load shared by the piles reduces
down to 72% of the total building loads. That raft depth
is equivalent three levels of basement which is not
uncommon in high-rise building design in Bangkok.
However, for shallower raft depth that the raft level is
still in the soft clay layer, piles have to carry more than
90% of building loads, thus the piled raft foundation
design concept does not offer much benefit.
－ 77 －
Fig. 14 Foundation configuration in this study
Amornfa, et al.
The 3 methods of analysis that are carried out are:
Method 1 (M1): Piled foundation analysis by the
plate on pile springs method. The analysis is made using
SAP2000: The method is most commonly used in current
Bangkok design practice. In this method the contribution
of soil below the raft, the pile-pile and pile-raft
interactions are ignored. The analysis model consists of
900 plate elements and 25 vertical springs as shown in
Fig. 15. In order for the results of this method of analysis
to be rightly compared with the results of 3D FEM
analysis, the pile spring stiffness is obtained from a 3D
FEM simulation of a single bored pile of the same size
and length installed in the same Bangkok subsoils as
used in the piled raft foundation analysis, Fig. 16. The
result shows a spring stiffness value of 7.22 x 105 kN/m,
which was equal to 1.06EA/L. The value is within the
range reported by Kiattivisanchai (2001), but lower than
the mean value of 2EA/L. For the plate on pile springs
analysis the use of lower value of pile spring stiff across
the board will not have effect on the load distribution
among piles and bending moment in the raft, it only
affect settlements of raft.
Fig. 15 The plate on pile springs method (M1)
Fig. 17 The plate on pile and soil springs method (M2)
ks 
Es
 Area
B(1   2 )
(6)
For the undrained modulus of the stiff clay (Es) of
40,500 kN/m2 and Poisson ratio of 0.495, the soil spring
stiffness (ks) was calculated as 5.36x104 kN/m. Although
this method of analysis considers the contribution of soil
resistance below the raft, the pile-pile interactions could
not be directly considered.
Method 3 (M3): 3D FEM piled raft foundation
analysis. This rigorous 3D analysis using PLAXIS 3D
Foundation program can directly consider the details of
subsoil layers within which the piles are installed and the
interactions among piles soil and raft. The soil and pile
models consist of 18,194 elements, mainly 15-node
wedge elements as shown in Fig. 18. Thin plate elements
were used to simulate the raft. The subsoil conditions
and engineering properties as well as the staged
simulation are the same as described in the early section.
Fig. 16 Load-settlement curve from simulation of
the pile load test by PLAXIS 3D Foundation
Method 2 (M2): Piled raft foundation analysis by the
plate on both pile and soil springs. The analysis is
similar to M1 except for that soil springs between piles
are included. In the analysis, the soil springs were
modeled at 1-m spacing underneath the entire raft. The
model consists of 900 plate elements and 25 pile springs
and 200 soil springs as shown in Fig. 17. The stiffness of
soil spring was obtained from the modulus of subgrade
equation proposed by Vesic (1961);
Fig. 18 The 3D FEM (M3)
Results of Load case 1: uniform load
The comparison on the results for Load Case 1 is
shown in Table 4. The result of the conventional method
is also included. According to the conventional method,
－ 78 －
Current practice on foundation design of high-rise buildings in Bangkok, Thailand
all piles in the raft will carry the same value of axial load
with FS of 2.50. Piles take 100% of the building load.
For the M1, the load distribution among piles is slightly
non-uniform. As expected, the corner and edge piles
carry a slightly higher loads than the inner and center
piles (in terms of FS: 2.37-2.48 as compared to 2.602.62). Again in M1, piles take 100% of building loads.
On the results of M3, the load sharing of piles reduces to
76.9%. The soil under the raft shares 23.1% of the
building loads. FS of corner and edge piles from M3 are
3.01-3.10 as compared to 3.50-3.52 for inner and center
piles that is significantly higher than the normal design
specification.
In term of bending moment in the raft which will
affect the required amount of steel rebar, M3 yields
much lower negative bending moment than that from M1
only 25% and 50% of M2 (Fig. 19). The positive
bending moment of the 3 methods is not so different.
This suggests that the rigorous pile raft foundation
analysis using 3D FEM can result in a more economical
design. In term of foundation settlement, M3 gives
higher settlement than M1 and M2, Fig. 20. The
calculated settlement at the center of raft from M3 is
highest (14.8 mm), while from M1 and M2 are only 5.9
and 3.8 mm, respectively. The magnitude is still small
(lower than 15 mm) for the foundation without reduced
number of piles from the plate on springs method.
Fig. 20 Raft settlement from various methods of
analysis: Uniform load case
Fig. 19 Bending moment in raft from various methods
of analysis: Uniform load case
Results of Load case 2: line and point load
The comparative summary of the results for this load
case is given in Table 5. For the piled raft foundation
analysis (M2 & M3) the results in term of load shared by
piles and load distribution among piles of the raft are of
a similar trend to that obtained from Load case 1,
although the magnitudes are slightly different. In term of
load distribution among piles, the application of this
non-uniform load case result in a much larger variation
in axial load acting on the piles from all M1, M2, and
M3 than the case of uniform load case, whereas the
conventional analysis shows the uniform pile load
Table 4 Load distribution on piles from various methods: Uniform load case
Method of analysis
Conventional
Piled foundation by Plate on pile
springs – M1
Piled raft foundation by Plate on both
pile and soil springs – M2
Piled raft foundation by 3D FEM – M3
Corner pile
4500 (2.50)
Load on pile, kN (Factor of safety)
Edge pile
Middle-strip pile Center pile
4500 (2.50)
4500 (2.50)
4500 (2.50)
% Load shared by
piles
100.00
4748 (2.37)
4527 (2.48)
4322 (2.60)
4296 (2.62)
100.00
2904 (3.87)
2825 (3.98)
2737 (4.11)
2735 (4.11)
62.37
3734 (3.01)
3624 (3.10)
3194 (3.52)
3218 (3.50)
76.87
－ 79 －
Amornfa, et al.
distribution. The load taken at the corner pile is 3-4
times larger than that transmitted to the piles in the
middle strip zone.
The bending moment and settlement of raft for the
load case are shown in Figs. 21 and 22, respectively. M3
gives the lowest negative and positive moment in raft.
The positive moment is largest at the center where a
large concentrated point load is applied but the negative
bending moment is largest over the corner pile that the
largest axial pile load occurs. On settlement, the
settlement at the center of raft from M3 was highest
(15.4 mm), while from the M1 and M2 are only 5.3 and
4.3 mm, respectively. The values are not so much
different from the uniform load case but the raft
curvature produced by the differential settlement is more.
The results of the comparative analysis above
indicate that real load distribution among piles
underneath the raft can be so much different from that
assumed in the conventional method of calculation. The
discrepancy is larger for the case of a more non-uniform
load from the superstructure of the building acting on the
raft. For the case of uniform load, the pile load
Fig. 21 Bending moment in raft from various methods
of analysis: Line and point load case
distribution from both the plate on springs method and
3D FEM analysis is somewhat uniform, i.e. all piles
carry similar level of axial load. In the case of a nonuniform building load over the raft for which a large
concentrated load as from a core locating at the raft
center, the piles at the corner carry much higher axial
load than the piles near the center of the raft. If the
selection of the single pile capacity in the design is based
on the results of the plate on springs type of analysis as
practiced by some designers, the design will be grossly
too conservative. For Bangkok subsoils, when the raft is
placed down to the depth of the stiff clay layer, the 3D
FEM shows that a portion of building loads that can be
carried by the soil underneath the raft may be around
30%, thus the determination of the number of piles on
the basis of conventional calculation is much too
conservative. The short term settlement of the raft as
calculated from the rigorous 3D FEM for a piled raft
foundation that the number of piles is determined from
the conventional method of calculation is in the order of
15 mm (uniform load case) which is within the normal
Fig. 22 Raft settlement from various methods of
analysis: Line and point load case
Table 5 Load distribution on piles from various methods: Line and point load case
Method of analysis
Conventional
Piled foundation by Plate on pile
springs – M1
Piled raft foundation by Plate on both
pile and soil springs – M2
Piled raft foundation by 3D FEM – M3
Corner pile
4500 (2.50)
Load on pile, kN (Factor of safety)
Edge pile
Middle-strip pile Center pile
4500 (2.50)
4500 (2.50)
4500 (2.50)
% Load shared
by piles
100.00
7471 (1.51)
4865 (2.31)
2351 (4.79)
3913 (2.88)
100.00
4613 (2.44)
3104 (3.62)
1419 (7.93)
3199 (3.52)
61.14
5218 (2.16)
3780 (2.98)
1584 (7.10)
4386 (2.56)
71.15
－ 80 －
Current practice on foundation design of high-rise buildings in Bangkok, Thailand
acceptable level of pile foundation performance used in
Bangkok. However, both methods of plate on pile
springs and plate on pile and soil springs show much
smaller settlement values (4-7 mm) which are not
realistic as the methods do not account for the interaction
between piles in carrying the vertical loads from the
superstructure.
Application of True Piled Raft Foundation Concept
In the comparative analysis made above, the number
of piles of the raft is not determined in accordance with
the real piled raft foundation design concept, but instead
according to the piled foundation design concept. The
analysis shows that when the raft is placed on the first
Bangkok stiff clay the load shared by piles is only 70%.
According to the real piled raft foundation concept, the
number of piles can be reduced. In order to explore the
application of the true piled raft foundation concept for
Bangkok subsoils, further 3D FEM analysis is made for
the same base case of raft foundation considered above.
The uniform load case is considered. The number of
piles is reduced from 25 to 21 and 16 but the layout
pattern remains as a uniformly distributed spacing over
the raft area. The new layouts for 21 and 16 pile numbers
are shown in Fig. 23. Table 6 shows the load distribution
on piles between the 25-, 21-, and 16-pile cases (pile
group FS of 2.50, 2.10 and 1.60, respectively). Normally,
FS of piles in a piled raft foundation design can be
accepted as low as 1.50 in order to allow soil underneath
the raft to carry an optimum load sharing level. For an
economical piled raft foundation design alternative, it
was suggested that piles may be loaded to close to their
ultimate capacity. This concept was described by Poulos
Fig. 23 Pile layout for raft on 25, 21 piles and 16 piles
(2001). Piles may be designed at lower FS Using this
concept, average settlement may be increased. But for
piled raft foundation, differential settlement can be
controlled. Therefore, it is likely to be considerably more
economical than the design with conventional FS (2.5).
The results of the analysis show that the load shared
by piles of the 21-pile and 16-pile cases only slightly
increase from that of the 25-pile case. However, the axial
load carried by each pile of the group significantly
increases. By expressing in term of FS of each pile, the
minimum value for 21-pile case is 2.40, reduced from
3.01 for 25-pile case, and it is 1.86 for 16-pile case. The
use of reduced number of piles in the raft has a negative
effect on the induced bending moment in the raft as well
as the settlement. However, the increase in the
magnitude of settlement is so significant (increased from
14.8 mm to 21.3 mm for 25-pile to 16-pile cases), Fig.
24. It has been reported that for piled raft foundation
design the magnitude of raft settlement of 100 mm was
acceptable when differential settlement can be controlled
(Poulos 2001). In term of bending moment, the effect
can be seen in Fig. 25. The maximum positive moment
increases by about 100% at some locations and the
negative moment increases by about 25%.
To illustrate economic advantage of the true piled
raft foundation design concept in Bangkok, a cost
comparison of a foundation of a 30-storey high-rise
building is demonstrated. The building has useable floor
area of 36,000 m2. Consequently, total applied load on
Fig. 24 Comparison on raft settlement between 25-, 21and 16-pile cases
－ 81 －
Amornfa, et al.
Table 7. In the comparison, the 513 million Baht is the
cost of building superstructure, basement and raft
excluding piles. The cost of pile is 140,000 Baht per pile.
The quantity of steel rebar in raft is approximately at
219,800 Bath/m3. Based on change of moment, steel
rebar in raft is assumed to be increased by 50 and 100%
for 151- and 93-pile cases, respectively. While using the
piled raft foundation concept with low FS, the total cost
is reduced by 9.5 million Baht or 1.75%. This cost
comparison example is a simple demonstration. The cost
changing due to load share and moment changing may
not always be straight forward. These behaviors should
be iteratively determined using rigorous 3D FEM.
CONCLUSIONS
Current design practice of foundation of high-rise
buildings in Bangkok and rooms for improvement are:
- From interview survey with one-third of design
companies of high-rise buildings in Bangkok, it was
found that most of the design works are handled by
structural engineers. Geotechnical specialists are
engaged in some design firms for detailed analysis of
piled foundations.
- Usually, in-depth 3D piled foundation analysis is
not made. The number of piles is mostly determined by
the conventional combined stress equation. And most
common method of detailed analysis is the plate on pile
springs analysis. Piled foundation design concept is
usually used in which piles are considered to carry all
building loads even for the case that raft in the stiff clay
is used as a pile cap. Thus the current design practice is
conservative and do not lead to economical design.
- The piled raft foundation design concept is
proposed for improvement of the foundation design of
high-rise buildings with basements extending to the stiff
Fig. 25 Bending moment in raft between 25-, 21- and
16-pile cases
foundation including the weight of the raft is 792,000 kN.
The ultimate capacity of a single bored pile was assumed
as 10,250 kN. According to the conventional design
calculation using FS of 2.5, the allowable pile capacity is
4,200 kN. Thus required number of piles is 189 piles.
If applying the piled raft foundation design concept,
the load shared by piles may be reduced to 80%.
Consequently, the number of piles is reduced to 151.
And using the allowable FS of 1.5, the allowable
capacity will be 6,830 kN, thus the number of piles can
be only 93 which was only half of pile number
determined by the conventional design. The cost
comparison of these design alternatives is summarized in
Table 6 Comparison on load distribution on piles between 25-, 21- and 16-pile cases
Analysis case
25 piles
21 piles
16 piles
Corner pile
3734 (3.01)
4693 (2.40)
4956 (2.27)
Load on pile, kN (Factor of safety)
Edge pile
Middle-strip pile
3624 (3.10)
3194 (3.52)
4274 (2.63)
4004 (2.81)
5618 (2.00)
-
Center pile
3218 (3.50)
3791 (2.97)
6063 (1.86)
% Load shared by piles
76.87
79.51
79.13
Table 7 Cost comparison between piled foundation and piled raft foundation concept
Design concept
Piled foundation
with F.S. = 2.5
Piled raft foundation
with F.S. = 2.5
Piled raft foundation
with F.S. = 1.5
Cost of Superstructure
+ Raft
No. of
pile
(Million Baht)
Cost of
Pile
Cost of Steel
rebar in Raft
Total cost
(Million Baht)
(Million Baht)
(Million Baht)
(Million Baht)
% of total cost
Reduced cost
513
189
26.5
4.0
543.5
0.0
0.00
513
151
21.1
6.0
540.1
3.4
0.63
513
93
13.0
8.0
534.0
9.5
1.75
－ 82 －
Current practice on foundation design of high-rise buildings in Bangkok, Thailand
clay layer. The rigorous 3D FEM shows that a load
shared by soil underneath the raft can reach 20-30% of
total building loads. If a reduction of FS of piles of 1.50
is used instead of 2.50 as currently used, the number of
piles in the raft can be reduced up to 50% and the load
shared by piles still remains around 70%. The analysis
shows that settlement would increase around 50% which
is not significant. The piled raft foundation design
concept is an economical method. It will also help
solving problem with the necessary use of large number
of piles at close spacing for high-rise buildings
constructed in small piece of land.
REFERENCES
Bowles, J.E. (1986). Mat design. Journal ACI, 83(6):
1010-1017.
Boonyarak, T. (2002). Behavior of polymer slurry for
wet-process bored pile construction in Bangkok
subsoils. Master thesis, Chulalongkorn University,
Bangkok.
Brinkgreve, R.B.J. and Broere, W. (2004). PLAXIS 3D
Foundation manual - version1. Balkema Publishers,
Netherlands.
Chaiyawan, P. (2006). Study of large bored pile capacity
behavior in Bangkok subsoil. Master thesis, Kasetsart
University. Bangkok.
de Sanctis, L. and Mandolini, A. (2006). Bearing
capacity of piled rafts on soft clay soils. Journal of
Geotechnical and Geoenvironmental Engineering,
132(12): 1600-1610.
de Silva, N. S. K. N. (1980). Analysis of bored piled in
Bangkok clay by finite element method. Master
thesis, Asian Institute of Technology, Bangkok.
Fernando, G. S. K. (1992). Load-settlement analysis of
bored piles in Bangkok subsoils using the finite
element method. Master thesis, Asian Institute of
Technology, Bangkok.
Katzenbach, R., Arslan, U. and Moormann, C. (2000).
Piled raft foundation in Germany. In H. J.A. (Ed.),
Design applications of raft foundations. London,
Thomas Telford Ltd: 323-391.
Kiattivisanchai, S. (2001). Evaluation of seismic
performance of an existing medium-rise reinforced
concrete frame building in Bangkok. Master thesis,
Asian Institute of Technology, Bangkok.
Kitiyodom, P. and Matsumoto, T. (2002). A simplified
analysis method for piled raft and pile group
foundations with batter piles. International Journal
for Numerical and Analytical Methods in
Geomechanics, 26: 1349-1369.
Ng, K. C. (1983). The construction problems and
performance of large bored piles in second sand layer.
Master thesis, Asian Institute of Technology,
Bangkok.
Oonchittikul, S. (1990). Performance of bored pile in
Bangkok subsoils. Master thesis, Asian Institute of
Technology, Bangkok.
Phienwej, N., Giao, P.H. and Nutalaya, P. (2006). Land
subsidence in Bangkok, Thailand. Engineering
Geology, 82: 187-201.
Pimpasugdi, S. (1989). Performance of bored, driven and
auger press piles in Bangkok subsoil. Master thesis,
Asian Institute of Technology, Bangkok.
Poulos, H. G. (2001). Piled raft foundations: design and
application. Geotechnique, 51(No.2): 95-113.
Poulos, H. G. and Davis, E. H. (1980). Pile foundation
analysis and design. New York: Wiley.
Randolph, M. F. (1983). Design of pile raft foundations.
Recent Development in Laboratory and Field Tests
and Analysis of Geotechnical Problems, Bangkok:
525-537.
SEAFCO public company limited (2011). SEAFCO
brochure.
Retrieved
April
2011,
from
http://www.seafco.co.th
Soontornsiri, A. (1995). Behavior and performance of
grouted bored pile in Bangkok subsoils. Master thesis,
Asian Institute of Technology, Bangkok.
Submaneewong, C. (1999). Behavior of instrumented
barrette and bored piled in Bangkok subsoils. Master
thesis, Chulalongkorn University, Bangkok.
Suwanwiwattana, P. (2002). Bangkok subsoil properties
by GRASS-GIS. Master Thesis, Kasetsart University,
Bangkok.
Teng, W.C. (1962). Foundation design. Prentice-Hall,
Inc. New Jersey.
Thasnanipan, N., Aye, Z. Z., Sunmaneewong, C. and
Boonyarak, T. (2006). Application of top-down
construction method for deep excavations in urban
area of Bangkok. International Symposium on
Underground Excavation and Tunnelling. Bangkok.
Vesic, A. S. (1961). "Beam on elastic subgrade and the
Winkler’s hypothesis" 5th ICSMFE,Vol.1: 855-850.
Wikimedia foundation. (2011) Skycraper. (in Thai).
Retrieved March 2011, from http://th.wikipedia.org/
Yamashita, K., Kakurai, M. and Yamada, T. (1994).
Investigation of a piled raft foundation on stiff clay.
Proceeding of the 13th International Conference on
Soil Mechanics and Foundation Engineering, New
Delhi: 543-546.
－ 83 －